News

Dragon capsule (via SpaceX)

SpaceX (Space Exploration Technologies Corp.) has recently announced that they will launch have launched their Dragon free-flying reusable space craft on Saturday March 19 at 1:15 AM Pacific time. This will mark the first time in history that a commercial company has launched a manned space vehicle into space to rendezvous with the International Space Station.

The Dragon re-usable space craft was designed using three main components which feature a nosecone that’s used as a shield during lift-off and houses the docking adapter needed to connect to external hatches found on the ISS. The second component featured is the spacecraft itself and was designed to be configured based on payload specifications and houses the avionics, RCS (Reaction Control System or thruster control systems) system, parachutes and other un-pressurized cargo/systems. The third component featured on the spacecraft is the Trunk which is used for un-pressurized cargo, solar arrays and thermal radiators needed to power the craft.

The re-usable vehicle will be launched atop of SpaceX’s Falcon 9 two-stage heavy-lift rocket which uses 10 Merlin 1C liquid oxygen and kerosene motors (9 on the first stage and 1 on the second). The mission will bring much needed supplies to the ISS as well as challenge the Dragon in a series of tests designed to test the feasibility of using commercial craft for future missions contracted through NASA and other organizations. If all goes well the launch will take place at Cape Canaveral at 4:15 AM Eastern Time and is expected to return a few hundred miles off the coast of California two weeks later. For those interested in watching the launch head over to SpaceX.com which will start broadcasting 40 minutes prior to launch.

Cabe

http://twitter.com/Cabe_e14

ARM Cortex A-9 (via TSMC)

Today's processor industry is largely controlled by two companies, Intel and ARM Holdings. Intel produces processors running in most of today’s laptops, desktops, and servers. On the other hand, ARM largely dominates the quickly growing mobile industry. Both are looking to invade each others markets soon by developing processors with high performance and low power consumption, or a strong performance per watt ratio.

Taiwan Semiconductor Manufacturing Company (TSMC) may have just made ARM Holdings future in the CPU market a bit more promising. At TSMC, they have recently ran a 28nm dual-core ARM Cortex A-9 processor at a max speed of 3.1GHz. The clock speed is 55% higher than present and is about twice as fast as its 40nm counterpart at TSMC. Additionally, the ARM chips also have the advantage of very little heat dissipation, giving them the ability to be densely packed together with one another.

Two of ARM's many partners include Nvidia and Calxeda, are both looking to produce ARM based processors to compete with Intel. Calxeda is working on producing chips for servers that work more efficiently. Such as implementing overlapping operations during each clock cycle to allow better speed handling. The method gives them an efficiency boost and may work to an advantage for large data retrieving applications such as web hosting.

TSMC also produces mobile chips for Nvidia. The successful high speed processing test can also mean good things for Nvidia. Nvidia is working on a custom ARM based processor to use in desktops and laptops to compete with Intel. The CPU project dubbed Project Denver has already been in development for some time, but TSMC's latest breakthrough could give the project a large boost.

The coming products produced from this “competition” should give us some interesting products in the future. Both companies will not easily be letting other companies invade their markets. The server market is worth $50 billion due to the rise of cloud computing and use of social networking. In addition, everyone can see the rise in the uses of tablets and smart phones.  The competition will lead us into the future of processor technology, which will be developed with these two companies paving the way.

Cabe

http://twitter.com/Cabe_e14

A tear-jerking introduction of the technology

A cerebrally controlled robotic system is being developed by a team of researchers from Brown University, Harvard Medical School, Massachusetts General Hospital and a host of others could give paralyzed people the ability to use robotic limbs to manipulate objects for themselves. Called ‘BrainGate’, the brain-controlled system allows the user to control a robotic limb through thought. To do this, the team implants a wireless microelectrode array (Neural Interface System) at 4 X 4mm directly on the motor cortex portion of the brain that controls motor function. The series of electrodes (100 in all) on the chip pick up the brain's activity associated with arm movement and sends the signals to a series of computers that use software (unknown at this time) to decode the brains activity. The computers then translate those signals into a series of instructions that tell a robotic arm to move and grasp an object based on the user’s desired intentions. The researchers are presently using two types of robotic arms, which are being continuously developed by DLR Institute of Robotics and Mechatronics and DEKA Research and Development Corporation.

DLR robotic hand/arm concept

The bigger of the two robotic arms being used by the researchers is DLR’s Hand Arm System, which is an external robotic arm made for more robust applications where impacts with heavy objects are nonconsequential (factory and warehouse work?).  The arm consists of a series of mechatronic compliance actuators with 52 drives and over 100 position sensors. The units hand alone features 38 individual tendons with each connected to an individual motor to provide tension and stiffness. The fingers use a similar configuration that uses two separate motors for individual grasping and tension based on the object being manipulated. The arm is so robust that you can actually beat it with a baseball bat without damaging any of the electronics.

Deka arm system

The second arm that the team is working with is DEKA Research and Development’s ‘Luke’ Arm (named after Luke Skywalker's mechanical hand). The arm is actually a robotic prosthesis that was designed for amputee patients and was developed as a DARPA tetraplegia project. The titanium Arm was designed to be roughly the same size as a typical human appendage and houses all of its electronics, motors and actuators inside (exactly how and what technology was used is currently unknown). The prosthesis features 18 degrees of movement which was accomplished by using rigid-to-flex circuit boards that were folded into ‘origami’ shapes placed inside the titanium housing. A vibrational motor at the top of the arm lets the user know how much pressure is needed to grasp an object through varying degrees vibration depending if the wearer is holding an egg or a brick.

Cabe

http://twitter.com/Cabe_e14

(All images and video courtesy of Crown Institute for Brain Science)

(Left) Erin Treacy Solovey wearing the Brainput device (Right) Artistic concept (via MIT & Erin Solovey)

When it comes to multitasking we as humans try the best we can. While we all have a modicum of ability, some are better than others. It suffices to say, we could all use a boost to become more efficient in our multitude of multitasking efforts, which is why a team of researchers has developed an unconventional solution to the problem. Led by Erin Solovey from MIT’s Humans and Automation Lab, the team has designed a system called ‘Brainput’ that can off-load some of our brains multitasking skills to a computer which is way more efficient at doing multiple things than we could ever hope to be. They system uses a portable low-cost version of a functional magnetic resonance imager called ‘fNIRS’ (functional Near-Infrared Spectroscopy) to measure the activity going on in the brain. The measurements are monitored and processed (using two probes) in real-time using Boxy software (from ISS). The information is then analyzed by a software engine (created using both Matlab and Weka tools) to look for specific patterns associated when the individual is multitasking. When the system has learned these patterns the software kicks in and is able to help the user with the task at hand.

A maze was created to test Brainputs effectiveness where a subject had to navigate through using two robots simultaneously. The operator using the fNIRS system was constantly switching back and forth between them and once the software learned the patterns it was able to engage sensors in the robots to help the user with their guidance. While the robots were autonomous, the test subject’s performance did indeed improve. While Brainput is still in its early development stages, it could be implemented into many applications in the future like helping you drive while you’re momentarily distracted or used during surgery with robotic assistance. What if the system could be used wirelessly? If you have an automated laundry machine, you could be slaving at work and washing your laundry at home at the same time! The possibilities are endless.

Cabe

http://twitter.com/Cabe_e14